The neuroleptic threshold hypothesis states that the lowest neuroleptic dose which produces a clinically detectable increase in hypokinesia-rigidity affords an individual schizophrenic patient all the therapeutic antipsychotic benefit available to that patient; and that higher doses provide no additional therapeutic benefit but may result in increased extrapyramidal side effects and subjective distress. Our pilot work suggests that the haloperidol doses at which acutely psychotic schizophrenic patients first develop clinically detectable evidence of hypokinesia-rigidity are distributed around the mean of 4.2 plus-minus 2.4 mg daily. We propose to openly adjust the daily haloperidol doses of newly admitted acute schizophrenic patients to each patients' individually determined neuroleptic threshold dose during the first 10 days of treatment, and to maintain patients at this (now fixed-) dose in an open trial for two weeks. Pilot data suggest that over 70% of patients will respond clinically. In order to determine whether improvement continues at the neuroleptic threshold dose, or whether a dose increase above neuroleptic threshold provides greater improvement (or any improvement for nonresponders), responding and non-responding patients will be randomly assigned (double-blind) to an additional two weeks trial at either the neuroleptic threshold dose or a higher dose (10 or 15 mg haloperidol daily).